“It was like the Titanic”-Some thoughts on the social amplification of risk

Certain risk events attract a lot of media coverage and this can have knock-on effects. Recent prime time TV coverage of the stricken cruise ship Viking Sky, battered by heavy seas with rescue helicopters circling overhead, will understandably have put some people off booking an arctic cruise for their next holiday! The Norwegian Tourist Board and the Cruise Line companies alike may experience some loss of visitors, revenues and perhaps falling stocks.

The Social Amplification of Risk Framework, or SARF¹, attempts to come to terms with such vexing matters of public concern, where there is genuine danger but the public response to the risk is said to be exacerbated by the media. Examples are all too frequent. A recent article in Risk Analysis² applies the framework to an outbreak of Zika virus³ . Recently people worried about whether risk surrounding the toxic chemical Fipronil⁴ found in some Dutch eggs was exaggerated. As time goes by, the real risk behind these headline grabbers can seem undeserving of all the fuss. People then accuse the media of whipping-up such controversies, fuelling public concern and sometimes forcing the government into knee-jerk reactions. A picture is often painted of a hype-crazed media intent upon over-dramatizing relatively unimportant risks. SARF captures this disquiet, defining broadcasters and newspapers, along with social media as social stations of amplification. To express this idea in more general terms, SARF characterises controversial risk events in terms of some sort of problematic, variable, mediated, societal response. Psychology Today recently summarized this popular notion, claiming that amplified risk information creates public hysteria⁵.

Diagram: Representation of the social amplification of risk, after Kasperson and colleagues

Louder or Riskier?
The plausibility of this view is easy to understand, but here my intention is to issue a warning to approach this framework with care. From the outset, questions over what is supposedly being amplified have dogged SARF. Is the media coverage itself somehow disproportionate to the risk, or is the risk itself materially increased: is it louder or riskier? I offer to chart a course through these troubled waters.

Louder
SARF is built upon on an electronic/acoustic metaphor of amplification, conjuring up the idea of increasing loudness and playing to a common-sense understanding of the way risk information is experienced. ‘Volume effects’, as described by SARF pioneer Ortwin Renn⁶ , are said to intensify (amplify) messages about risks via societal communication processes depicted as signals and receivers.

The first problem is that while loudness may be appropriate as an everyday heuristic, risk analysis needs to be more exacting; it needs to be clear about what is being amplified and how. Clarity is required because one of the defining features of risk analysis is an underlying objective to mitigate risk. In other words, risk is characterised primarily so that it can be managed. This is where a loudness reading of SARF quickly starts to unravel, as first highlighted by the sociologist Are Rip in his article, ‘Should Social Amplification of Risk Be Counteracted’⁷ . Rip wants to know whether SARF is making value judgements or describing objective miscalculation. In short, how do we know whether a particular risk is being amplified or not? Whilst I, or anyone else can say that a risk is attracting too much noise, this is just opinion. To demonstrate ‘too much’ from an analytical perspective we might expect a corresponding value that is not excessive. SARF offers no benchmark or any sort of scale that helps to make the loudness concept useable. Without any objective mechanism SARF cannot help in determining how much anyone ought to worry about anything.

Gauging appropriate levels of media reporting is similarly bereft of any sound foundation. Even a methodologically appropriate media analysis (a common way for SARF to look for evidence of amplification), such as counting the word ‘Zika’ on Twitter, needs some mechanism to decide how many ‘Zikas’ are too many. To their credit, the authors of the Zika paper fully acknowledge this shortcoming.

Attenuation
Adding to the confusion, SARF introduces the flip side of amplification to describe risks that receive too little attention. Unfortunately, without a useable scale, preferably with a helpful mid-point between amplification and attenuation, or benchmark that risk management efforts might aspire to, analysts only have tools which are based upon subjectivity to decide whether anything is being amplified or attenuated. This is not straightforward. Unsurprisingly, in most controversial cases there is disagreement about proportionality. From marine incidents to disease outbreaks, with fracking, Fipronil and Brexit along the way, experts and lay persons alike are typically divided. Some say we need to raise awareness of such issues; others say the risk is being exaggerated; and, perhaps one or two others who are in the middle will view the current response as about right. Agreement is rare, and may only emerge after the event when disaster has either struck or been avoided. Soberingly, we see evidence of coming together following the Grenfell Tower disaster, around which a consensus has emerged retrospectively, that the risk had gone under the radar.

I should point out that SARF has been used to reflect on social processes involving perceptions of risk and the role of the media. Such matters are rarely presented in purely quantitative terms and it would be wrong to think about SARF in terms of objective miscalculations or misrepresentations. While a cruise ship operator might legitimately point to low probabilities of serious incidents, there is no objective sense in which I can be wrong if I say, ‘the government should do more’, or even histrionically ‘it was like the Titanic!’ People are inevitably going to worry as much or as little as they please, and risk appetites are notoriously wide-ranging regardless of scientific understanding. My partner, who is a statistician, is a more nervous flyer than I am. Does her more sophisticated understanding of the relevant probabilities make her wrong and a risk amplifier, or am I wrong and a naively optimistic attenuator of aviation risk?

Beyond these variations between individuals in levels of anxiety, many societal risk controversies, which SARF seeks to explain, essentially have a political dimension. How much ought to be done about marine safety, Zika, Fipronil tainted eggs, fracking or GM food, are of course all questions that can benefit from robust scientific evidence. Yet in the real world they will all, like other issues demanding public attention, ultimately be subject to political priorities. The anthropologist Mary Douglas best articulates this view in terms of the ‘cultural theory of risk’⁸ . More recently, Andy Stirling⁹ (always well worth reading) takes an equally political perspective on risk. These authors get to the heart of the matter. Regarding political controversy as disproportionate to objective risk can, if we’re not careful, privilege techno-scientific institutional interests over legitimate, democratic freedoms. To label a risk amplified is, in practice, to stigmatise the framing; conversely, the way to promote more risk mitigation for an issue is to claim that it is going under the radar, that it is not attracting the appropriate attention, or in SARF terms, that it is being attenuated.

Riskier
Alongside ‘louder’, SARF ambivalently argues that the process of social amplification makes risks ‘riskier’. Amplification, according to SARF, occurs when specific risk characteristics spawn secondary and tertiary risks, typically in other sectors of society. For example, falling luxury cruise ship bookings may have far-reaching effects on ship-building, or the fortunes of coastal resorts. The Framework expresses this through an additional metaphor of ripples in a pond. Ripple effects might include political fall-out – an unfortunate Government Minister might have to resign.

Here, SARF faces the immediate problem that the amplification metaphor is ill-fitted. While most people readily understand interconnectedness – ‘for want of a nail the shoe was lost’ (as the poem goes), this isn’t what amplification brings to mind. The utility of the metaphor weakens further at the other end of SARF’s spectrum. If amplification means ‘riskier’, its flip side, attenuation, also seems to mean riskier since risks that go under the radar are clearly also consequential in terms of tertiary effects. Many disasters have been compounded because concerns were ignored or played-down. Viewing Grenfell Tower as ‘less risky’ while it was not in the public eye appears both a perverse and disturbing framing of this recent tragedy. In short, the amplification metaphor simply isn’t flexible enough to describe the complexity of real-world risk controversies.

Of course, some risks have more repercussions than others, but this is not generally down to the newsworthiness of corresponding event characteristics, as SARF would have it. In many risk episodes the major disruptor seems to be uncertainty about its likelihood or magnitude. This is unsurprising considering that risk is often defined as uncertainty about likelihood or magnitude. In the face of uncertainty about the scale or nature of a risk, experts often simply ‘get it wrong’. In many instances it is not the characteristics of an event about which the shouting was too loud, but rather what we did not know that proved most consequential to the trajectory of the crisis. We can’t argue coherently that the uncertain magnitude of a risk is a characteristic of the risk event without floundering into a tautology.

Louder and Riskier
So why not let SARF have it both ways, portraying risk as ‘louder’ and ‘riskier’ at the same time. In practice, mixing the electronic metaphor of signals and receivers with the naturalistic metaphor of stones in ponds seems to be where most SARF-based studies end up. But, rather like mamma in my favourite JJ Cale song, I don’t much like the band “all playing at the same time”¹⁰ . For me this muddling of metaphors, first pointed out by the sociologist Steve Rayner¹¹ , brings me back to my earlier objection. Almost anything can be said to be more consequential given more publicity. Glasgow Rangers Football Club will sell more replica shirts in Ulaanbaatar if they have more media presence. This in turn may help them to buy a new player and so on and so forth. It is not a unique feature of risk that media noise is materially consequential. To be blunt, the generalization that an event’s characteristics relate to the amount of press it receives is as ubiquitous and trivial as the idea that events typically have both characteristics and consequences. But even this trivial claim, that louder risks are riskier, is undermined because SARF also proposes that quieter risks can be riskier too. This is all loosely assembled without any mechanism to assess volume.

******

For me, dressing up SARF’s various truisms as a quasi-empirical framework for conducting risk analysis remains as problematic today as it did when Are Rip and Steve Rayner raised their metaphorical eyebrows 30 or so years ago. I can, however, understand the continuing appeal of SARF. It was created as a trans-paradigmatic bridge, which is no mean feat of engineering! It attempts to reconcile profoundly different perspectives on risk, notably psychology and anthropology.

The central danger has always been the temptation SARF provides to normatively juxtapose mediated popular opinion with expert judgement in a prejudiced way. As a risk researcher, I am delighted to see people being paid to carry out expert assessments of risk and conduct robust science. Where empirical evidence is available it must be given proper weight in management and policy. However, it is not for me or anyone else to tell passengers who likened their Norwegian cruise to the Titanic (Daily Mail)¹² that they are over-reacting, nor to decide that the press is over-egging the pudding by leading bulletins with distressing images. Ultimately, I’m loath to judge whether the material consequences of this or any other incident is disproportionate to an objective assessment of the real risk without a compelling reason, particularly in the face of uncertainty.

As the bridge of SARF continues to creak and sway, with researchers sometimes acknowledging its limitations, but often seemingly oblivious to them, SARF is able to do what it said on the tin! That is, to offer a conceptual framework, which was not originally advanced with the pretensions of a theory. My hope is that researchers crossing this venerable old structure will tread carefully.

References

¹Kasperson, R. E., O. Renn, P. Slovic, H. S. Brown, J. Emel, R. Goble, J. X. Kasperson and S. Ratick (1988). "The social amplification of risk: a conceptual framework." Risk Analysis 8(2): 177-187.
²https://onlinelibrary.wiley.com/journal/15396924
³Wirz, C. D., M. A. Xenos, D. Brossard, D. Scheufele, J. H. Chung and L. Massarani (xx date? xx)"Rethinking Social Amplification of Risk: Social Media and Zika in Three Languages." Ibid. 0(0).
⁴LINK to my previous blog post Eggs Benelux
⁵https://www.psychologytoday.com/us/blog/statistical-life/201806/society-is-megaphone-risk
⁶https://www.iass-potsdam.de/en/people/ortwin-renn (Renn’s homepage)
⁷Rip, A. (1988). "Should Social Amplification of Risk Be Counteracted?" Risk analysis 8(2): 193-197.
⁸Douglas, M. and A. Wildavsky (1982). Risk and culture : An essay on the selection of technological and environmental dangers. Berkeley, CA, University of California Press.
⁹Andy Stirling’s ResearchGate https://www.researchgate.net/profile/Andy_Stirling
¹⁰https://www.youtube.com/watch?v=88tA5zql6CQ
¹¹Rayner, S. (1988). "Muddling through metaphors to maturity: a commentary on Kasperson et al., The social amplification of risk." Risk Analysis 8(2): 201-204.
¹²https://www.youtube.com/watch?v=88tA5zql6CQ
 

Disclaimer: The views expressed in this blog post are the views of the author(s), and not an official position of the institute or funder.